Exercise 2_1_1: Working with OSM Data

1

Pick a city of the world to your liking and retrieve its bounding box. Then plot it: What do you see?

Clues

The function you may search is part of the osmdata package.

solution

dresden <-
  osmdata::getbb("Dresden")

plot(dresden)

# We see two points, which build the extremes of the bounding box.

2

Please choose one or two building types you are interested in and set them as key and value pairs. You can find a list of building types in the Overpass API documentation. But don’t forget to set the timeout query parameters first using the osmdata::opq() function.

Clues

First, you specify the bounding box like before, second, the query parameters, and third, the key and value pairs. Try two use a pipe workflow as it makes it a bit easier.

solution

dresden <-
  osmdata::getbb("Dresden") %>% 
  osmdata::opq(timeout = 25*100) %>% 
  osmdata::add_osm_feature(
    key = "building", 
    value = c("house", "residential")
  )

3

Download the data using the osmadata::osmdata_sf() function and extract only the polygons.

Clues

The downloaded data are a list. The polygons are a named list element that you can extract with its name osm_polygons, just like a variable in a data table.

solution

dresden <-
  osmdata::getbb("Dresden") %>% 
  osmdata::opq(timeout = 25*100) %>% 
  osmdata::add_osm_feature(
    key = "building", 
    value = c("house", "residential")
  ) %>% 
  osmdata::osmdata_sf() %>% 
  .$osm_polygons

4

Take some time to browse through the data. Depending on the building type you have chosen, you may find some interesting information. You can also plot the data you have just downloaded.

Clues

You may consider converting the data into a tibble using tibble::as_tibble() and maybe a sf::st_as_sf() afterwards for a nicer browsing experience.

solution

library(tmap)

dresden <-
  osmdata::getbb("Dresden") %>% 
  osmdata::opq(timeout = 25*100) %>% 
  osmdata::add_osm_feature(
    key = "building", 
    value = c("house", "residential")
  ) %>% 
  osmdata::osmdata_sf() %>% 
  .$osm_polygons %>% 
  tibble::as_tibble() %>% 
  sf::st_as_sf()

dresden

## Simple feature collection with 14557 features and 122 fields
## Geometry type: POLYGON
## Dimension:     XY
## Bounding box:  xmin: 13.57913 ymin: 50.97482 xmax: 13.96545 ymax: 51.17778
## Geodetic CRS:  WGS 84
## # A tibble: 14,557 × 123
##    osm_id   name  X3dr.type abandoned abandoned.man_made access addr.city addr.country addr.housename
##    <chr>    <chr> <chr>     <chr>     <chr>              <chr>  <chr>     <chr>        <chr>         
##  1 22986774 <NA>  <NA>      <NA>      <NA>               <NA>   Dresden   <NA>         <NA>          
##  2 22986800 <NA>  <NA>      <NA>      <NA>               <NA>   Dresden   DE           <NA>          
##  3 22986802 <NA>  <NA>      <NA>      <NA>               <NA>   Dresden   DE           <NA>          
##  4 22986803 <NA>  <NA>      <NA>      <NA>               <NA>   Dresden   DE           <NA>          
##  5 22986804 <NA>  <NA>      <NA>      <NA>               <NA>   Dresden   DE           <NA>          
##  6 22986805 <NA>  <NA>      <NA>      <NA>               <NA>   Dresden   DE           <NA>          
##  7 22986806 <NA>  <NA>      <NA>      <NA>               <NA>   Dresden   DE           <NA>          
##  8 22986815 <NA>  <NA>      <NA>      <NA>               <NA>   Dresden   DE           <NA>          
##  9 22986816 <NA>  <NA>      <NA>      <NA>               <NA>   Dresden   DE           <NA>          
## 10 22986940 <NA>  <NA>      <NA>      <NA>               <NA>   Dresden   <NA>         <NA>          
## # … with 14,547 more rows, and 114 more variables: addr.housenumber <chr>, addr.postcode <chr>,
## #   addr.street <chr>, addr.suburb <chr>, addr.suburb.en <chr>, alt_name <chr>, amenity <chr>,
## #   architect <chr>, area <chr>, attraction <chr>, brand <chr>, brand.wikidata <chr>, building <chr>,
## #   building.architecture <chr>, building.colour <chr>, building.levelPlan <chr>, building.levels <chr>,
## #   building.levels.underground <chr>, building.material <chr>, building.part <chr>,
## #   building.roof <chr>, building.roof.colour <chr>, building.roof.shape <chr>, building.use <chr>,
## #   castle_type <chr>, check_date <chr>, cn_tud.token <chr>, colour <chr>, comment <chr>, …

tm_shape(dresden) +
  tm_polygons()